Invariant Policy Optimization: Towards Stronger Generalization in Reinforcement Learning

Keyphrases

• Generalization in Reinforcement Learning 100%
• Policy Optimization 100%
• Physical Properties 50%
• Learning Algorithm 50%
• Reinforcement Learning 50%
• Fundamental Challenges 50%
• Invariance Principle 50%
• Linear Quadratic Regulator 50%
• Generalization Performance 50%
• Successful Actions 50%
• Grid World 50%
• World Problems 50%
• Operating Domain 50%
• Open Doors 50%
• Policy Gradient Methods 50%
• Unseen Domain 50%

Computer Science

• Reinforcement Learning 100%
• Policy Optimization 100%
• Learning Algorithm 50%
• Gradient Method 50%
• Generalization Performance 50%
• Robotics 50%